At the end of LD22, I was quite proud of myself. I had made a game, for the first time, and exceeded all my expectations. I had to deal with a library that I learned about just a couple of days before, almost no preparation, and lots of weird problems packaging the software. So it came naturally to me that this time out I would take all the lessons of LD22, build on top of them, and make an awesome game. Oh, how wrong I was!
But I’m getting ahead of myself. I went into great lengths to prepare myself for this LD. I participated in the warm up, and build myself a neat little shooter for the 10 hours that I allowed myself to spend in the warm up. Then I bought all kinds of food that I could or could not need, since food was another thing that I felt was lacking in my first LD. Since the theme would be revealed 10AM my time, I slept early, and planned to go to a restaurant near my home to think about the theme away from the computer.
All was working well so far. “Tiny World” was not my favorite in the list of themes, but I did not dislike it. The inspiration came when, at the restaurant, the expresso machine broke. I immediately remembered all those children stories about small gnomes fixing househood items, hidden from the adults. Specially, I was thinking about the recent Ghibli movie, Arrietty, which I liked a lot. So the idea started to form on my head. A family of tiny creatures, trying to live unnoticed in a human’s house.
I spent about 20 minutes thinking about a princess-maker style management game, where you managed a family of “borrowers”, sent them on missions to explore, or expand, or other things, and tried to keep your profile low. But then I thought that a management game was not a good call for LD — management games are not everyone’s couple of tea, and they can be quite hard to balance. I decided to change the focus a bit, in order to make something more likeable, and I came with the idea of controlling a single of these creatures, in an action/stealth game.
Maybe that was my mistake, maybe I should have gone with the management game I wanted to make… anyways.
The idea for the game was this — a creature living underground, had to forage the above world for items, food, treasures. All while avoiding the humans and other hazards in the big world above, and using the spoils to improve their own tiny world. It was still a good idea on paper. But when I started laying out the plan for the game engine, I panicked. I realized that I needed creatures of widely different sizes, on a complex world, and I had no idea of how to handle collision. In my previous game, collision was a non-issue, and I was getting scared that I would spend a LOT of time trying to learn how to make a quad tree or something like that. It did not make any sense, I was just panicking, and in panicking I lost a lot of time.
Eventually I decided to re-implement a tile based engine just like in my previous game, with three different sizes of sprites: 1, 9 and 25 tiled units. I implemented a simple and dirty routine for square collision detection on those sizes. Then I started to work on some early lose conditions, on the idea that I needed to have a prototype with a win/lose condition as soon as possible. So I implemented foraging for food, dying of starvation, and the cat predator.
But by then, it had already been over 30 hours, and I had very little time to do much else.
In the end, my game idea about living among humans and exploring, fixing their stuff, and building your base while remaining undetected became a silly game about stealing food from cats to give to your fat queen . The game didn’t even have any humans.
Other than the end result, not all was lost, though. The preparation certainly paid off: I did not have any problems regarding my programming environment (no mysterious bugs) or the living environment. Submitting, publishing, time-lapsing all worked without a hitch. I got a lot of practice with the Gimp, and learned to use some tools in different ways. I also learned how to create and add music to the game without requiring a lot of composition knowledge. Autotracker is a wonderful little tool that I learned about too late into the compo, but will certainly be a center piece for future dares. In fact, I might go as far as create a bunch of random tunes to just listen to them while playing DF or something like that. Also, I managed to insert a lot of little neat widgets into the game that made me a bit proud of myself, such as a generic handler for mobs using inheritance, a text message listener that shapes the message to an arbitrary window, dealing with real time, a pause screen, etc. Too bad that these neat tricks can’t replace an actual game :-/
Well, there is always next time! I still feel pumped for programming, so maybe I can get some more game making experience before LD24.
I finished my game, but it fell way short of my expectations, couldn’t implement critical features.
* My preparation was top notch: I wanted for nothing, biological or technical
* I managed to solve many technical problems I had in my LD22 game, such as music and message box formatting.
* I panicked early in the dare because of lack of familiarity with game programming, lost a lot of time.
* I misjudged the time I needed, critical features and playtesting were missing from the game.
* If I like a niche theme or a game style, I should just go for it. With 1000 participants, LD is not a popularity contest.
* Working with multiple small levels uses the time better than working with one complex level.
* Autotracker rocks.